An articulated robot is provided in a conveying part that performs loading/unloading of a workpiece to/from a processing part in a processing apparatus for continuously processing workpieces in a plurality of processing parts which are successively installed. In a semiconductor processing apparatus which is typical for such a processing apparatus, a cluster type processing apparatus in which processing parts as a working part are radially disposed around a conveying part, and an in-line type processing apparatus in which processing parts are linearly disposed are used. A cluster type processing apparatus has a problem in that the number of the processing parts are limited to 5 or 6 at maximum. In contrast to this, an in-line type processing apparatus has an advantage that the number of the processing parts can be increased as needed.
As a conventional articulated robot, there is disclosed a robot arm apparatus including: a first arm which is rotatably mounted to a base via a first transmission mechanism; a second arm which is rotatably mounted to a distal end part of the first arm via a second transmission mechanism; and a fork member which is rotatably linked to a distal end part of the second arm via a third transmission mechanism and for mounting an object to be transferred (for example, Patent Literature 1). The fork member has a distal end part which is bifurcated, and a base end part which is supported by the second arm. In the above described Patent Literature 1, the robot arm apparatus is configured such that the fork member always advances from its distal end part into a processing part.